Design Thinking
What it is
Design Thinking is a human-centered approach to innovation that uses methods from design practice to match people's needs with what is technologically feasible and commercially viable. It is not a step-by-step process but a set of mindsets and methods that can be applied at different PDLC stages.
The most widely used model is the Double Diamond (British Design Council, 2004), which describes two phases of divergent and convergent thinking:
- Diamond 1 — Problem space: Diverge (explore many problems) → Converge (define the right problem)
- Diamond 2 — Solution space: Diverge (explore many solutions) → Converge (validate the right solution)
Design Thinking addresses desirability risk — "Do people actually need and want this?" — which is the most common cause of product failure.
Authoritative sources (external)
| Resource | Executive summary (why it's linked here) |
|---|---|
| IDEO — Design Thinking | Origin framework from IDEO — the firm that popularized Design Thinking. Defines the five modes (Empathize, Define, Ideate, Prototype, Test) and human-centered design principles. |
| Stanford d.school | Academic anchor for Design Thinking education — process guides, teaching materials, and the "bootcamp bootleg" (free methodology cards). |
| British Design Council — Double Diamond | Double Diamond model — the canonical visualization of diverge/converge across problem and solution spaces. |
| Nielsen Norman Group — Design Thinking | Practitioner summary with evidence-based guidance on when Design Thinking works (and when it doesn't). |
Core structure
The Double Diamond
graph LR
subgraph diamond1 ["Diamond 1: Problem Space"]
discover["Discover (diverge)"]
define["Define (converge)"]
end
subgraph diamond2 ["Diamond 2: Solution Space"]
develop["Develop (diverge)"]
deliver["Deliver (converge)"]
end
discover -->|"many problems explored"| define
define -->|"right problem defined"| develop
develop -->|"many solutions explored"| deliver
deliver -->|"right solution validated"| buildPhase["SDLC A-F"]
Five modes (IDEO / Stanford d.school)
| Mode | Diamond | Activity | Output |
|---|---|---|---|
| Empathize | D1 — Diverge | Observe and engage with users; understand their experiences, needs, and motivations | Empathy maps, interview notes, observation logs |
| Define | D1 — Converge | Synthesize research into a clear problem statement (point of view) | Problem statement, "How Might We" questions, persona refinements |
| Ideate | D2 — Diverge | Generate a broad range of solution ideas; defer judgment; build on ideas | Idea clusters, sketches, concept cards, prioritization matrix |
| Prototype | D2 — Converge (early) | Build quick, cheap representations of ideas to test with users | Paper prototypes, wireframes, clickable mocks, concierge MVPs |
| Test | D2 — Converge (late) | Put prototypes in front of users; gather feedback; refine or pivot | Usability test results, validation evidence, iteration notes |
Mapping to PDLC phases
Design Thinking maps most strongly to PDLC P1–P2 (pre-build discovery and validation):
| PDLC phase | Design Thinking mode(s) | Activities |
|---|---|---|
| P1 Discover Problem | Empathize + Define | User research, empathy mapping, problem framing, "How Might We" synthesis |
| P2 Validate Solution | Ideate + Prototype + Test | Brainstorming, rapid prototyping, usability testing, concept validation |
| P3 Strategize | Outputs inform strategy | Validated problem statement and solution concept feed vision, OKRs, business case |
| SDLC C Design | Prototype (refined) | Higher-fidelity design artifacts inform engineering design phase |
| P5 Grow | Empathize + Test (continuous) | Post-launch user research and testing for iteration |
When to use Design Thinking
| Situation | Fit |
|---|---|
| New product / new market | Strong — high uncertainty about user needs demands empathy-first approach |
| Complex problem, unclear solution | Strong — diverge-converge helps explore before committing |
| Feature improvement on mature product | Moderate — Empathize mode verifies assumptions; faster iteration than full Double Diamond |
| Technical / infrastructure work | Weak — users are internal; needs are more knowable; skip to specification |
| Bug fix / maintenance | Skip — problem is already defined by the bug itself |
Anti-patterns
| Anti-pattern | Fix |
|---|---|
| Design Thinking as a workshop, not a practice | A one-day "Design Thinking workshop" is not product discovery. Embed the modes into weekly team practice. |
| Skipping Empathize | Jumping to solutions without understanding users is the #1 failure mode. Require evidence from real users, not stakeholder assumptions. |
| Diverge without converge | Infinite ideation without commitment. Set clear converge criteria: "By Friday we choose 3 ideas to prototype." |
| Prototype fidelity too high | Low-fi prototypes (paper, wireframes) are faster and cheaper for early testing. Save high-fi for validation, not exploration. |
| Design Thinking without delivery | Beautiful insights and prototypes that never become products. Bridge to SDLC by setting clear gates (PDLC G2) after Test mode. |
Further reading
- Lean Startup — Complementary: adds business viability testing to Design Thinking's desirability focus
- Opportunity Solution Trees — Visual structure for organizing the Ideate mode
- PDLC.md — P1 and P2 — How these modes map to PDLC phases with exit criteria
- PDLC-SDLC Bridge — Full lifecycle context